title: MODIFYING METACOGNITIVE STRATEGY INSTRUCTION BASED COGNITIVE ACADEMIC LANGUAGE LEARNING APPROACH (CALLA) FOR WRITING CLASS IN RELATION TO STUDENTS’ LEVEL OF CRITICAL THINKING creator: MELINA SARI, 1523042017 subject: L Education (General) = Pendidikan subject: PE English description: This study aimed at investigating the effect of modified CALLA on students’ writing achievement and the use of metacognitive strategies, finding out the effect of students’ critical thinking level on their writing achievement, and discovering the process of implementing modified CALLA in the language classroom. This research was experimental research based on one−group pretest−posttest design. Science classes (XI IPA 1 and XI IPA 2) were taken as the subject of the research by using purposive random sampling. There were three classifications consisting of: 17 low critical thinker, 31 middle critical thinker, and 4 high critical thinker. There were four instruments administered in this study, namely: Academic Potency Test (APT), Metacognitive Strategy Use in Writing (MSUW), Writing Test, and observation. The researcher collected the data by conducting three stages of activities. They were distributing Academic Potency Test (APT), Metacognitive Strategy Use in Writing (MSUW) and Writing Test, conducting treatment and employing observation sheet, administering Metacognitive Strategy Use in Writing (MSUW), and Writing Test. In analyzing the data, the researcher used Statistical Package for Social Science (SPSS) version 21 program. The data obtained from Metacognitive Strategy Use in Writing (MSUW) and writing test were calculated before and after treatment. Repeated Measures t-test was used to compare the means score from the result of pretest and posttest. Since this study dealt with the effect of students’ critical thinking level on their writing achievement, Univariate Analysis of Variance (One Way Anova) was used to see whether there was correlation between students’ critical thinking level toward their writing achievement. The significance level was in 0.05 even the hypothesis was approved if sign